Assignment2_Kyrkos_Konstantinos.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Assignment II: Exploring Why Civil Resistance Works\n",
    "\n",
    "---\n",
    "\n",
    ">Konstantinos Kyrkos <br>\n",
    ">Registration Number p3351907 <br>\n",
    ">kkyrkos@aueb.gr"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This notebook provides my solutions for the second assignment for the Practical Data Science class at Data Science MSc at AUEB. <br>\n",
    "Bollow you will find the solution to all 7 exercises. <br>\n",
    "The only prerequisite for running the commands below is to provide the already given WCRWreplication.dta file which contains the data to which the assignment is related to.\n",
    "\n",
    "## 1: Frequency of Nonviolent and Violent Campaign End Years\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import statsmodels.api as sm\n",
    "import statsmodels.formula.api as smf\n",
    "import statsmodels.graphics as sgr\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "from matplotlib.collections import LineCollection\n",
    "from statsmodels.stats.outliers_influence import summary_table\n",
    "import math\n",
    "\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "   defect  outcome  peakmembership  eyear  byear            location  \\\n",
      "0     1.0        5         50000.0   1959   1956                Cuba   \n",
      "1     0.0        3          5000.0   1920   1918               Haiti   \n",
      "2     1.0        5         30000.0   1985   1985               Haiti   \n",
      "3     0.0        3           650.0   1924   1916  Dominican Republic   \n",
      "4     1.0        3             NaN   1965   1965  Dominican Republic   \n",
      "\n",
      "               campaign                target  lccode  tccode  ...   lmtnest  \\\n",
      "0      Cuban Revolution        Batista regime    40.0    40.0  ...  1.694107   \n",
      "1           Caco Revolt       U.S. occupation    41.0     2.0  ...       NaN   \n",
      "2                        Jean Claude Duvalier    41.0    41.0  ...  2.797281   \n",
      "3  Dominican insurgency       U.S. occupation    42.0     2.0  ...       NaN   \n",
      "4              leftists       Loyalist regime    42.0    42.0  ...  2.856470   \n",
      "\n",
      "       ef  loclpolity1  loclcap1  loclcap2  loclcap3  loclcap4     mempc  \\\n",
      "0  0.2134         -9.0       0.0       1.0       0.0       0.0  1.226839   \n",
      "1     NaN          NaN       NaN       NaN       NaN       NaN       NaN   \n",
      "2  0.0950         -9.0       1.0       0.0       0.0       0.0  1.200712   \n",
      "3     NaN          NaN       1.0       0.0       0.0       0.0       NaN   \n",
      "4  0.3868          0.0       1.0       0.0       0.0       0.0       NaN   \n",
      "\n",
      "   demdum  recur1  \n",
      "0     0.0     0.0  \n",
      "1     0.0     NaN  \n",
      "2     0.0     0.0  \n",
      "3     0.0     NaN  \n",
      "4     0.0     0.0  \n",
      "\n",
      "[5 rows x 85 columns]\n"
     ]
    }
   ],
   "source": [
    "data = pd.read_stata(\"WCRWreplication.dta\")\n",
    "\n",
    "print(data.head(5))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We want to visualize the frequency of nonviolent and violent campaigns. <br> \n",
    "The frequency of the campaigns should be counted based on their end years, at decades. Also we want the violent and nonviolent campaign counts to be stacked on each other. <br>\n",
    "For that we start by getting our data to the form we want. <br>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "years = data.groupby(data.eyear // 10 * 10).count().eyear\n",
    "eyears = years.index\n",
    "\n",
    "non_violent = data[data.nonviol==1].groupby(data.eyear // 10 * 10).count().eyear\n",
    "non_violent[1900]=0\n",
    "non_violent.sort_index(inplace=True)\n",
    "\n",
    "violent = data[data.viol==1].groupby(data.eyear // 10 * 10).count().eyear"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now that we have our data in the preferred form, we simply draw a stackplot to present our data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "labels = [ \"Violent\",\"Non-Violent \"]\n",
    "\n",
    "fig, ax = plt.subplots()\n",
    "ax.grid(True)\n",
    "ax.stackplot(eyears, violent, non_violent, labels=labels, edgecolors='white', linewidths=.9, )\n",
    "ax.set_axisbelow(True)\n",
    "plt.xlabel('decade')\n",
    "plt.ylabel('campaigns')\n",
    "plt.margins(x=0)\n",
    "plt.legend(loc=\"upper left\")\n",
    "fig = plt.gcf()\n",
    "fig.set_size_inches(20,10, forward = False)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2: Number of Nonviolent Campaigns and Percentage of Success"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "At this point we want to create a plot showing, for each decade from 1940 onwards, the number of nonviolent campaigns and the percentage of success. Our plot will have two vertical axes, one for each of the metrics. <br>\n",
    "First we have to manipulate our data in order to prepare them for our plot."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "non_violent_from_40s = data[data.nonviol==1].groupby(data.eyear // 10 * 10).count().eyear[3:]\n",
    "\n",
    "successful = data[(data.nonviol==1) & (data.success==1)].groupby(data.eyear // 10 * 10).count().eyear[2:]\n",
    "\n",
    "success_perc = successful / non_violent_from_40s"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Next, we draw our plot with the data we have gathered."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x720 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "success_df = pd.DataFrame({'number of non violent campaigns':non_violent_from_40s, 'success perc':success_perc})\n",
    "success_df.plot(kind= 'bar' , secondary_y= 'success perc' , rot= 0 , figsize=(20, 10))\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3: Success Rate by Decade"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For this exercise we want to create a plot showing the success rate for violent and nonviolent campaigns by decade. <br> \n",
    "Our plot should show that nonviolent campaigns have higher success probabilities like the book of well-received book of Erica Chenoweth (\"Why Civil Resistance Works\") suggests."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "First again we start by preparing our data for the plot which will come later."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "non_violent_from_40s = data[data.nonviol==1].groupby(data.eyear // 10 * 10).count().eyear[3:]\n",
    "\n",
    "nonviol_successful = data[(data.nonviol==1) & (data.success==1)].groupby(data.eyear // 10 * 10).count().eyear[2:]\n",
    "\n",
    "nonviol_success_perc = nonviol_successful / non_violent_from_40s\n",
    "\n",
    "violent_from_40s = data[data.viol==1].groupby(data.eyear // 10 * 10).count().eyear[4:]\n",
    "\n",
    "viol_successful = data[(data.viol==1) & (data.success==1)].groupby(data.eyear // 10 * 10).count().eyear[4:]\n",
    "\n",
    "viol_success_perc = viol_successful / violent_from_40s\n",
    "\n",
    "years = data.groupby(data.eyear // 10 * 10).count().eyear\n",
    "eyears = years.index[4:]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now that we have our data ready for plot, we can present our results."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure()\n",
    "ax = fig.add_subplot() \n",
    "ax.set_axisbelow(True)\n",
    "N = len(eyears)\n",
    "ind = np.arange(N) \n",
    "width = 0.4     \n",
    "plt.grid(True)\n",
    "plt.bar(ind, nonviol_success_perc, width, label='nonviolent')\n",
    "plt.bar(ind + width, viol_success_perc, width, label='violent', color = 'red')\n",
    "plt.xlabel('decade')\n",
    "plt.ylabel('success rate')\n",
    "plt.title('Campaign type and success rate over decades')\n",
    "fig = plt.gcf()\n",
    "fig.set_size_inches(20,10, forward = False)\n",
    "plt.xticks(ind + width / 2, eyears)\n",
    "plt.legend(loc='upper left')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 4. Twenty-Five Largest Resistance Campaigns, 1900-2006"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For this exercise we want to create a table with the twenty-five largest resistance campaigns, for all years, based on their number of members at the highest point (peak membership). <br>\n",
    "The table will include the membership, the start year, the end year, the location, the target, whether it was violent or not, and whether it was successful or not. <br>\n",
    "Finally we will try to check if we can notice anything about Greece."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We start by collecting and sorting our data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "data_ordered_by_membership = data.sort_values(by=['peakmembership'],ascending=False).head(25).reset_index()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Next we choose the preferred columns and make our new table more readable/presentable removing duplicate columns."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "    peakmembership  byear  eyear        location               target violent  \\\n",
      "0        4500000.0   1937   1945           China  Japanese occupation     yes   \n",
      "1        2000000.0   1977   1979            Iran    Shah Reza Pahlavi      no   \n",
      "2        2000000.0   1986   1983     Philippines     Ferdinand Marcos      no   \n",
      "3        1000000.0   1988   1988           Burma       military junta      no   \n",
      "4        1000000.0   1989   1989           China     Communist regime      no   \n",
      "5        1000000.0   1967   1968           China         anti-Maoists     yes   \n",
      "6        1000000.0   1922   1949           China    Kuomintang regime     yes   \n",
      "7        1000000.0   2006   2006          Mexico      Calderon regime      no   \n",
      "8        1000000.0   2005   2005         Lebanon        Syrian forces      no   \n",
      "9        1000000.0   1993   1999         Nigeria        Military rule      no   \n",
      "10       1000000.0   1984   1985          Brazil        Military rule      no   \n",
      "11        700000.0   1990   1991          Russia            Anti-coup      no   \n",
      "12        700000.0   1983   1989           Chile     Augusto Pinochet      no   \n",
      "13        550000.0   1956   1957           China     Communist regime      no   \n",
      "14        500000.0   1989   1989  Czechoslovakia     Communist regime      no   \n",
      "15        500000.0   2002   2003      Madagascar     Radsiraka regime      no   \n",
      "16        500000.0   1963   1963          Greece    Karamanlis regime      no   \n",
      "17        500000.0   2001   2004         Ukraine        Kuchma regime      no   \n",
      "18        500000.0   2001   2001     Philippines       Estrada regime      no   \n",
      "19        400000.0   1941   1945            USSR      Nazi occupation     yes   \n",
      "20        400000.0   1991   1993      Madagascar     Didier Radsiraka      no   \n",
      "21        400000.0   1953   1953    East Germany     Communist regime      no   \n",
      "22        340000.0   1958   1975         Vietnam      U.S. occupation     yes   \n",
      "23        300000.0   1943   1945           Italy      Nazi occupation     yes   \n",
      "24        300000.0   1944   1944          Poland    German occupation     yes   \n",
      "\n",
      "   successful  \n",
      "0          no  \n",
      "1         yes  \n",
      "2         yes  \n",
      "3          no  \n",
      "4          no  \n",
      "5          no  \n",
      "6         yes  \n",
      "7          no  \n",
      "8         yes  \n",
      "9         yes  \n",
      "10        yes  \n",
      "11        yes  \n",
      "12        yes  \n",
      "13         no  \n",
      "14        yes  \n",
      "15        yes  \n",
      "16        yes  \n",
      "17        yes  \n",
      "18        yes  \n",
      "19         no  \n",
      "20        yes  \n",
      "21         no  \n",
      "22        yes  \n",
      "23         no  \n",
      "24         no  \n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\ipykernel_launcher.py:3: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  This is separate from the ipykernel package so we can avoid doing imports until\n",
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\ipykernel_launcher.py:4: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  after removing the cwd from sys.path.\n"
     ]
    }
   ],
   "source": [
    "largest_campaigns = data_ordered_by_membership[['peakmembership','byear','eyear','location','target',\n",
    "                                                'viol', 'nonviol', 'success', 'failure']]\n",
    "largest_campaigns['violent'] = np.where(largest_campaigns.viol==1,'yes','no')\n",
    "largest_campaigns['successful'] = np.where(largest_campaigns.success==1,'yes','no')\n",
    "del largest_campaigns['viol']\n",
    "del largest_campaigns['nonviol']\n",
    "del largest_campaigns['success']\n",
    "del largest_campaigns['failure']\n",
    "print(largest_campaigns)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "After we have presented our data, we check if we can find anything about Greece and we do in the 17th row from above. <br>\n",
    "We print this row again in order to comment on the data we found."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "peakmembership               500000\n",
      "byear                          1963\n",
      "eyear                          1963\n",
      "location                     Greece\n",
      "target            Karamanlis regime\n",
      "violent                          no\n",
      "successful                      yes\n",
      "Name: 16, dtype: object\n"
     ]
    }
   ],
   "source": [
    "print(largest_campaigns.loc[16])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can see that there is an entry for Greece for the events that occurred in 1963. <br>\n",
    "The campaign happened at the day of the funeral of G. Lambrakis. <br>\n",
    "Lambrakis being a peacemaker of his time and a member of the left wing party was killed by members of the shadow government at the time when K. Karamanlis was the prime minister of Greece, ruling with excessive use of violence. <br>\n",
    "This event led the people to indignation and made them want a fresh start, ending Karamanlis's authoritarian government. <br>\n",
    "In other words, this nonviolent campaign which unfortunately started from a tragic death of a person, ended succeeding its cause."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 5. The Effect of Participation on the Probability of Campaign Success"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For the purpose of this exercise we want to examine the effect of participation on the probability of campaign success. \n",
    "We will run a logistic regression for success on membership per capita, controlling for the location’s logged population the year the campaign ends. <br>\n",
    "After running the logistic regression we will explain our results. <br>\n",
    "Having done that we will create a scatter plot showing the logged participants per capita, on the $x$ axis, and the probability of success, on the $y$ axis."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "First we prepare our data and run the logistic regression for success on membership per capita, controlling for the location's logged population the year the campaign ends."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Optimization terminated successfully.\n",
      "         Current function value: 0.594790\n",
      "         Iterations 5\n",
      "                           Logit Regression Results                           \n",
      "==============================================================================\n",
      "Dep. Variable:                success   No. Observations:                  163\n",
      "Model:                          Logit   Df Residuals:                      160\n",
      "Method:                           MLE   Df Model:                            2\n",
      "Date:                Sun, 01 Dec 2019   Pseudo R-squ.:                  0.1294\n",
      "Time:                        11:20:59   Log-Likelihood:                -96.951\n",
      "converged:                       True   LL-Null:                       -111.35\n",
      "Covariance Type:            nonrobust   LLR p-value:                 5.551e-07\n",
      "==============================================================================\n",
      "                 coef    std err          z      P>|z|      [0.025      0.975]\n",
      "------------------------------------------------------------------------------\n",
      "Intercept     -5.8199      1.921     -3.030      0.002      -9.584      -2.055\n",
      "mempc          4.4892      0.986      4.554      0.000       2.557       6.421\n",
      "lnpop          0.0907      0.129      0.706      0.480      -0.161       0.343\n",
      "==============================================================================\n",
      "                  0         1\n",
      "Intercept -9.584271 -2.055451\n",
      "mempc      2.557314  6.421034\n",
      "lnpop     -0.161198  0.342643\n"
     ]
    }
   ],
   "source": [
    "from statsmodels.formula.api import logit\n",
    "\n",
    "success_mod = logit(\"success ~ mempc + lnpop\", data).fit()\n",
    "\n",
    "print(success_mod.summary())\n",
    "print(success_mod.conf_int())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Explaining and interpreting our results from above we can see the following. <br>\n",
    "<ul>\n",
    "  <li>The Pseudo R-squared value of 0.129 is fairly low. This is only logical due to the fact that we only used two of our many variables to express/predict the value of the success variable and so we have only expressed only 10% of the success variable.</li>\n",
    "  <li>Each one-unit change in mempc will increase the log odds of getting success by 4.4892</li>  \n",
    "  <li>Each one-unit change in lnpop will increase the log odds of getting success by 0.0907</li>  \n",
    "  <li>The intercept is the log odds of having success when all the independent variables are held at zero; that is  $e^{-5.8199} \\approx 0.0029679$. That means that $p / (1-p) = 0.0029679$</li>\n",
    "   <li>It is also interesting to observe that the P-value of mempc is really low (almost zero) meaning that this variable is really significant for all levels of significance</li>\n",
    "</ul>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now we create a scatter plot showing the logged participants per capita, on the $x$ axis, and the probability of success, on the $y$ axis."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1296x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.scatter(data.mempc, success_mod.predict(data))\n",
    "plt.xlabel('Participants per capita, logged')\n",
    "plt.ylabel('Probability of Success')\n",
    "plt.grid()\n",
    "plt.rc('axes', axisbelow=True)\n",
    "fig = plt.gcf()\n",
    "fig.set_size_inches(18,10, forward = False)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 6. The Level of Participation Tipping Point"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For this exercise we want to use the results we obtained in the previous question, to find the percentage of the population that is the tipping point for success in a campaign.\n",
    "We start by collecting the data we need and sorting them to see our results. <br>\n",
    "First we start by setting up again the data we will use for this exercise."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Optimization terminated successfully.\n",
      "         Current function value: 0.594790\n",
      "         Iterations 5\n",
      "                           Logit Regression Results                           \n",
      "==============================================================================\n",
      "Dep. Variable:                success   No. Observations:                  163\n",
      "Model:                          Logit   Df Residuals:                      160\n",
      "Method:                           MLE   Df Model:                            2\n",
      "Date:                Sun, 01 Dec 2019   Pseudo R-squ.:                  0.1294\n",
      "Time:                        11:21:04   Log-Likelihood:                -96.951\n",
      "converged:                       True   LL-Null:                       -111.35\n",
      "Covariance Type:            nonrobust   LLR p-value:                 5.551e-07\n",
      "==============================================================================\n",
      "                 coef    std err          z      P>|z|      [0.025      0.975]\n",
      "------------------------------------------------------------------------------\n",
      "Intercept     -5.8199      1.921     -3.030      0.002      -9.584      -2.055\n",
      "mempc          4.4892      0.986      4.554      0.000       2.557       6.421\n",
      "lnpop          0.0907      0.129      0.706      0.480      -0.161       0.343\n",
      "==============================================================================\n",
      "                  0         1\n",
      "Intercept -9.584271 -2.055451\n",
      "mempc      2.557314  6.421034\n",
      "lnpop     -0.161198  0.342643\n"
     ]
    }
   ],
   "source": [
    "success_fit = logit(\"success ~ mempc + lnpop\", data).fit()\n",
    "predicted_success = success_fit.predict(data)\n",
    "print(success_mod.summary())\n",
    "print(success_mod.conf_int())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The first approach we try is also the simplest and we do it by creating a new dataframe with the predicted data, along with the data from the variables we used to make predictions. <br>\n",
    "Keeping only the values which correspond to success over 0.49 we should get the tipping point for success in a campaign. <br>\n",
    "Since our data are logged, we use exp to de-log them and then do the math to get our population percentage."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "        mempc      lnpop   lmembers   success\n",
      "262  1.020733  13.534897  13.815511  0.497576\n",
      "140  1.091469  11.554644  12.611538  0.532006\n",
      "269  1.092927  10.534025  11.512925  0.510537\n",
      "299  1.096068  10.037786  11.002100  0.502808\n",
      "216  1.110516   9.283026  10.308952  0.501905\n",
      "1.3239423300881805 corresponds to success:  0.4975759966393706\n",
      "2.622986376685759 corresponds to success:  0.5320059497025783\n"
     ]
    }
   ],
   "source": [
    "values = pd.DataFrame({'mempc':data.mempc, 'lnpop':data.lnpop, 'lmembers':data.lmembers, 'success':predicted_success})\n",
    "values = values[values.success>=0.49]\n",
    "values = values.sort_values(by='mempc',ascending=True)\n",
    "print(values.head())\n",
    "perc = np.exp( values.lmembers.tolist() ) / np.exp( values.lnpop.tolist() )\n",
    "perc.sort()\n",
    "print(perc[0], \"corresponds to success: \",values.iloc[0].success)\n",
    "print(perc[1], \"corresponds to success: \",values.iloc[1].success)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As we can see from the values above, there is a clear gap in our success values and so we cannot give a clear border for our tipping point. Nevertheless we can give an interval in which the tipping point of success happens. That interval located around 1.32 and 2.62 of the population."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Another way of looking at our problem is forming an equation from the coefficients we found. Then the tipping point would be the line of our equation. So we form our equation from the data we can find in the summary above.<br>\n",
    "We want sigmoid(4.4892 * mempc + 0.0907 * lnpop  -5.8199) >=0.5. We added a virtual sigmoid function because although we have our coefficients, the final line would follow a sigmoid shape as we know from the theory of logistic regression. <br>\n",
    "So the tipping point is exactly this line and so we can transform the inequality to an equality and so we have <br> sigmoid(4.4892 * mempc + 0.0907 * lnpop  -5.8199) = 0.5. <br>\n",
    "We know that mempc = lmembers/lnpop and also we have to raise all values to exponential in order to remove the log from our variables.<br>\n",
    "sigmoid(4.4892 * (lmembers/lnpop) + 0.0907 * lnpop  -5.8199) = 0.5. <br>\n",
    "This line expressed by the two varibles (lmembers, lnpop) described the border on which we get values of 0.5 and any value above this line gives us the desired outcome of over 0.5."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 7. Nonviolent Resistance, Target Regime Type, Geographical Location"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "1. We want to examine whether nonviolent resistance remains significant even if we control for the regime type. <br>\n",
    "So we run a logistic regression of success on the nonviolence indicator controlling for the POLITY IV score of campaign target, the log of peakmembership, and the location’s logged population the year the campaign ends."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Optimization terminated successfully.\n",
      "         Current function value: 0.567143\n",
      "         Iterations 6\n",
      "                           Logit Regression Results                           \n",
      "==============================================================================\n",
      "Dep. Variable:                success   No. Observations:                  141\n",
      "Model:                          Logit   Df Residuals:                      136\n",
      "Method:                           MLE   Df Model:                            4\n",
      "Date:                Sat, 30 Nov 2019   Pseudo R-squ.:                  0.1627\n",
      "Time:                        17:26:02   Log-Likelihood:                -79.967\n",
      "converged:                       True   LL-Null:                       -95.506\n",
      "Covariance Type:            nonrobust   LLR p-value:                 2.953e-06\n",
      "==============================================================================\n",
      "                 coef    std err          z      P>|z|      [0.025      0.975]\n",
      "------------------------------------------------------------------------------\n",
      "Intercept     -0.1085      1.413     -0.077      0.939      -2.878       2.661\n",
      "nonviol        0.9048      0.464      1.952      0.051      -0.004       1.813\n",
      "tpolity        0.0520      0.030      1.744      0.081      -0.006       0.110\n",
      "lmembers       0.3751      0.128      2.926      0.003       0.124       0.626\n",
      "lnpop         -0.4363      0.142     -3.067      0.002      -0.715      -0.157\n",
      "==============================================================================\n",
      "                  0         1\n",
      "Intercept -2.877881  2.660870\n",
      "nonviol   -0.003837  1.813408\n",
      "tpolity   -0.006442  0.110417\n",
      "lmembers   0.123877  0.626331\n",
      "lnpop     -0.715148 -0.157497\n"
     ]
    }
   ],
   "source": [
    "success_mod = logit(\"success ~ nonviol + tpolity + lmembers + lnpop\", data).fit()\n",
    "\n",
    "print(success_mod.summary())\n",
    "print(success_mod.conf_int())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Explaining and interpreting our results from above we can see the following. <br>\n",
    "<ul>\n",
    "  <li>The Pseudo R-squared value of 0.162 is again fairly low. This is only logical due to the fact that we only used a few of our many variables to express/predict the value of the success variable and so we have only expressed only about 15% of the success variable.</li>\n",
    "  <li>Each one-unit change in nonviol will increase the log odds of getting success by 0.9048</li>  \n",
    "  <li>Each one-unit change in tpolity  will increase the log odds of getting success by 0.0520</li>  \n",
    "  <li>Each one-unit change in lmembers will increase the log odds of getting success by 0.3751</li>  \n",
    "  <li>Each one-unit change in lnpop will decrease the log odds of getting success by -0.4363</li>  \n",
    "  <li>The intercept is the log odds of having success when all the independent variables are held at zero; that is $e^{-0.1085}$</li> <br>\n",
    " Commenting on the significance of variables we observe the following:\n",
    "  <li>Almost all variables have p-values below 0.05 and all of them have values below 0.1. That means that all of them are considered significant for a significance level of 0.1 and the majority of them for significance a level of 0.05</li>\n",
    "So finally we can say that nonviolent resistance remains significant even if we control for the regime type\n",
    "</ul>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "2. We also want to examine whether nonviolent resistance remains significant even if we control for geographical location of the campaign. <br>\n",
    "So we run a logistic regression of success on the nonviolence indicator controlling for the geographical location of the campaign. <br>\n",
    "We leave randomly one variable describing the location out of the regression. We do this because, since we have 6 variables and each describes a region, having all the others except one, the outcome for this region will be expressed as what's left from all the others and so there is no use of adding this variable too."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Optimization terminated successfully.\n",
      "         Current function value: 0.591713\n",
      "         Iterations 6\n",
      "                           Logit Regression Results                           \n",
      "==============================================================================\n",
      "Dep. Variable:                success   No. Observations:                  323\n",
      "Model:                          Logit   Df Residuals:                      316\n",
      "Method:                           MLE   Df Model:                            6\n",
      "Date:                Sat, 30 Nov 2019   Pseudo R-squ.:                 0.08320\n",
      "Time:                        17:40:45   Log-Likelihood:                -191.12\n",
      "converged:                       True   LL-Null:                       -208.47\n",
      "Covariance Type:            nonrobust   LLR p-value:                 4.950e-06\n",
      "==============================================================================\n",
      "                 coef    std err          z      P>|z|      [0.025      0.975]\n",
      "------------------------------------------------------------------------------\n",
      "Intercept     -1.7374      0.430     -4.042      0.000      -2.580      -0.895\n",
      "nonviol        1.2582      0.265      4.751      0.000       0.739       1.777\n",
      "americas       1.0604      0.518      2.048      0.041       0.046       2.075\n",
      "asia           0.2612      0.505      0.517      0.605      -0.729       1.252\n",
      "africa         1.0235      0.485      2.112      0.035       0.073       1.974\n",
      "europe         0.4650      0.540      0.861      0.389      -0.594       1.524\n",
      "fsu            0.7645      0.628      1.218      0.223      -0.466       1.995\n",
      "==============================================================================\n",
      "                  0         1\n",
      "Intercept -2.579958 -0.894853\n",
      "nonviol    0.739171  1.777138\n",
      "americas   0.045556  2.075314\n",
      "asia      -0.729361  1.251673\n",
      "africa     0.073492  1.973593\n",
      "europe    -0.593634  1.523717\n",
      "fsu       -0.466093  1.995115\n"
     ]
    }
   ],
   "source": [
    "success_mod = logit(\"success ~ nonviol + americas + asia + africa + europe + fsu\", data).fit()\n",
    "\n",
    "print(success_mod.summary())\n",
    "print(success_mod.conf_int())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Explaining and interpreting our results from above we can see the following. <br>\n",
    "<ul>\n",
    "  <li>The Pseudo R-squared value of 0.0832 is again fairly low. This is again logical due to the fact that we only used the variables of the geographical location to express/predict the value of the success variable and so we have only expressed only the 1% of the success variable.</li>\n",
    "  <li>Each one-unit change in nonviol will increase the log odds of getting success by 1.2582</li>  \n",
    "  <li>Each one-unit change in americas will increase the log odds of getting success by 1.0604</li>  \n",
    "  <li>Each one-unit change in asia will increase the log odds of getting success by 0.2612</li>  \n",
    "  <li>Each one-unit change in africa will increase the log odds of getting success by 1.0235</li>  \n",
    "  <li>Each one-unit change in europe will increase the log odds of getting success by europe</li>  \n",
    "  <li>Each one-unit change in fsu will increase the log odds of getting success by 0.7645</li>  \n",
    "  <li>The intercept is the log odds of having success when all the independent variables are held at zero; that is  $e^{-1.7374}$</li>\n",
    "  Commenting on the significance of variables we observe the following:\n",
    "  <li>The nonviol variable again is really significant on all levels of significance with value under 0.001 and almost near to zero.</li>\n",
    "So finally we can say that nonviolent resistance remains significant even if we control the geographical location of the campaign.\n",
    "</ul>"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}